“…Since Cell Ranger is widely used, we investigated the extent to which the Cell Ranger results are concordant with our workflow. We processed 20 datasets ( Supplementary Table 1), chosen to contain a range of reads depths (from 8,860,361 to 721,180,737 reads per sample and 2,243 to 201,952 reads per cell) and to represent scRNA-seq from a range of tissues and species ( Arabidopsis thaliania 18 , Caenorhabditis elegans 19 , Danio rerio 20 , Drosophila melanogaster 21 , Homo sapiens 22,23 , Mus musculus [24][25][26][27][28] , Rattus norvegicus 28 ). We found a high degree of concordance with respect to quality control metrics (Figure 2a-h, Supplementary Figure 3).…”